[LWN Logo]
[LWN.net]

Bringing you the latest news from the Linux World.
Dedicated to keeping Linux users up-to-date, with concise news for all interests


Sections:
 Main page
 Security
 Kernel
 Distributions
 Development
 Commerce
 Linux in the news
 Announcements
 Back page
All in one big page

Other LWN stuff:
 Daily Updates
 Calendar
 Linux Stocks Page
 Book reviews
 Penguin Gallery

 Archives/search
 Use LWN headlines
 Advertise here
 Contact us

Recent features:
- RMS Interview
- 2001 Timeline
- O'Reilly Open Source Conference
- OLS 2001
- GaŽl Duval
- Kernel Summit
- Singapore Linux Conference
- djbdns

Here is the permanent site for this page.

See also: last week's LWN.

Leading items and editorials


Unix sucks. Or, at least, that's what Miguel de Icaza had to say in his opening keynote speech at the Ottawa Linux Symposium. This talk was, as he put it, "designed to insult everybody," and it may well have succeeded. Certainly Miguel, who has always been an energetic speaker who is unafraid of speaking his mind, was in fine form at OLS.

[Miguel] According to Miguel, free software does not yet have much to be proud of. We have still not provided a complete free software system, which addresses the needs of most users. There's not much in the way of new ideas, everything has been inspired by Unix. And, as we have noted, Miguel thinks that Unix sucks. (If you are thinking that GNOME sucks, however, it's probably "because you are using Red Hat's version.")

Unix is said to suck because it has long since ceased to be a platform for innovation. Unix, says Miguel, is stagnant. Unix's problems come from its longstanding approach of not deciding policy. The kernel does not decide policy; neither does the C library, the X libraries, or the window system in general. The people who decided that X users could pick their own window manager created a situation where there were many, many window managers to choose from; "they were smoking crack."

The Unix approach of not deciding policy is "a defense system for hackers," since that way nobody has to take responsibility for a bad decision.

But the real problem, according to Miguel, is that the Unix approach does not lead to any sort of significant code reuse. A list of modern applications was presented; it included names like Netscape, Acrobat, FrameMaker, StarOffice, and others. The amount of code sharing between those applications was pointed out as being almost zero. About the only thing they share is the C library, which does not provide much. A number of obviously common services (printer configuration and use was a favorite example) had been completely reimplemented for each application. The result is duplicated work and inconsistent interfaces. Things should not be that way.

Miguel is an open admirer of how Microsoft does software development. Microsoft's applications are built from components, which are highly reusable. That not only helps Microsoft; it enables independent developers to make use of high-level components from Office, Internet Explorer, and other applications. Linux has no equivalent set of components, and suffers from a much lower level of code reuse.

The mission of the GNOME project, at this point, is to change that situation. Thus the whole "Bonobo" system for defining and sharing components in a language-independent way. If Miguel has his way, Linux will soon have a set of high-level components of its own, and will be well on its way toward developing a collection of modern, user-friendly applications.

And, in fact, this component collection is going to get a jump start because....

StarOffice is to be released under the GPL. Sun took a step toward truly free software this week with the announcement that StarOffice, acquired by Sun last year, is to be released under the GPL. This is an interesting donation by Sun, especially when one looks at what the plans for this code are.

But first, why is Sun releasing the code? Since it became freely downloadable, StarOffice, one can presume, has not generated a great deal of revenue. So Sun is not losing much by opening up the source license. But, more importantly, Sun is presumably hoping that this move will help to make StarOffice the dominant office suite on the Linux platform. If (1) Linux continues to make progress on the desktop, and (2) Microsoft is sufficiently slow in making Office available, the holder of the desktop office suite standard, if there is one, could be in a good position. If StarOffice ends up with a good number of desktop users, it will enable Sun to press forward with its web-based version which, of course, will require lots of expensive Sun servers.

Should the free software community be pleased with this contribution? Certainly it is hard to get too upset about more code being available. With luck, some of the easier "bugs and bloat" problems can be dealt with by the user community, and perhaps parts of StarOffice will be useful in other applications. Gnumeric, for example, may be able to benefit through the inclusion of some of StarOffice's spreadsheet code.

But StarOffice is a great monster of a program. It's "do everything in one place" philosophy goes against everything that a good Unix utility is supposed to be: compact, reliable, and easily interfaced with other tools. StarOffice brings its own file formats, it works poorly with other X tools, and generally expects to be a world unto itself. It can be very frustrating to work with.

But consider the other interesting part of the StarOffice release: the code is going to be reworked, integrated with Bonobo and GTK, and released as a set of reusable components. If you are a GNOME programmer looking to put together a powerful application, you will be able to pick and use any pieces of StarOffice that your application needs. The result is likely to be an explosion of new, user-centric applications for Linux.

StarOffice will also be reworked to use a set of open, XML-based file formats, which is also a cause for celebration.

It's worth keeping in mind as well that the free software community doesn't always deal well with a sudden infusion of corporate code. Mozilla, unfortunately, is a classic example. For all the promise that Mozilla has, two years later very few people are using its code. While nothing is certain, StarOffice could well take a similar path. It is, according to the announcement, the largest single piece of GPL code out there. Getting a handle on all that will be a challenge. Just because it is open source does not mean that it will be successful. The fact that Collab.Net is being brought in to manage the code is a good sign, however.

Sun appears to be trying to do this right. The use of the GPL means that StarOffice code can be integrated with other applications. But breaking the code into components and integrating into GNOME means that many applications can be built without having to use the code directly. In the future, it may not be StarOffice itself which affects the life of Linux users; instead, it may be the wealth of applications built from its pieces. It will be interesting to see how this goes.

For more from the Ottawa Linux Symposium, see Jon Corbet's reports from the show.

On to other interesting tidbits this week .... SCO may be purchased by Caldera, reports this ZDNet article. This is reminiscent of some predictions that one of the post-IPO Linux companies would pick up SGI. SCO has been around for almost twenty years and has a long track record. Now it might be purchased by a veritable "upstart". Nonetheless, this looks like a potential good match. SCO's emphasis on reseller channels matches Caldera's long-term philosophy and their support services would put Caldera on a much better footing in competition with Red Hat, particularly in the international markets.

Meanwhile, the rumored price, $70 million in stock, seems incredibly small given the prices for small Linux startups only a few months ago.

Complete rewrite planned for Perl 6. At the ongoing Perl Conference 4.0, a meeting of the perl5-porters list was held to discuss Perl's long-term future. As a result, Larry Wall and Nathan Torkington announced plans for a complete rewrite for Perl 6. Frustration has developed as Perl has become increasingly difficult to improve, extend, and embed. The core will be reimplemented, to improve speed, and they'll try to take the opportunity to remove "cruft" from the current implementation.

Simultaneously, they are trying to revamp the Perl development community as well. Personality issues caused problems on the perl5-porters list during the last development cycle -- changes will be implemented to try and change that. Perl will be moving from following Larry Wall's vision to becoming a community-driven development effort. That speaks for some pretty large potential growing pains. The alternative, though, would be to allow Perl development to stagnate.

Linux in the telephony market. At a presentation at a recent local LUG meeting, LWN editor Liz Coolbaugh was asked about Linux in the telephony arena. She had little to report at that time and therfore found this Upside article on the relicensing of over a million lines of code under the GPL in 1998. The code was from the now-defunct Ingate telephony company.

Apparently Ingate CEO Martin Clinton was introduced to the potential of free software by Rich Bodo, then working at VA Research. As a result, one of his last acts as CEO of Ingate was the release of much of their software under the GPL. Now Martin and Rich are co-founders of a new company poised to make use of the result. This is an excellent precedent for commercial companies; dumping years of software development efforts into a dust-bin is an incredible waste. Free software offers an alternative, which can provide future commercial potential out of commercial failures.

Inside this week's Linux Weekly News:

  • Security: Computer Security Insurance, OpenHack update.
  • Kernel: API changes, 2.4, and the release cycle.
  • Distributions: Immunix, Embedded Linux, more Linux plans in China.
  • Development: Documentation, Free Pascal, Galeon browser, mod-snake.
  • Commerce: CompTIA and Linux+ certification, Open Source Projects
  • Back page: Linux links, this week in Linux history, and letters to the editor
...plus the usual array of reports, updates, and announcements.

This Week's LWN was brought to you by:


July 20, 2000

 

Next: Security

 
Eklektix, Inc. Linux powered! Copyright © 2000 Eklektix, Inc., all rights reserved
Linux ® is a registered trademark of Linus Torvalds